Fire as a motor of rapid environmental degradation during the earliest peopling of Malta 7500 years ago

HAL Id: hal-02083745

https://hal-amu.archives-ouvertes.fr/hal-02083745

Submitted on 29 Mar 2019

HAL is a multi-disciplinary open access archive for the deposit and dissemination of sci- entific research documents, whether they are pub- lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.

Fire as a motor of rapid environmental degradation during the earliest peopling of Malta 7500 years ago

Nick Marriner, David Kaniewski, Timmy Gambin, Belinda Gambin, Boris Vannière, Christophe Morhange, Morteza Djamali, Tachikawa Kazuyo,

Vincent Robin, Damien Rius, et al.

To cite this version:

Nick Marriner, David Kaniewski, Timmy Gambin, Belinda Gambin, Boris Vannière, et al.. Fire as a motor of rapid environmental degradation during the earliest peopling of Malta 7500 years ago.

Quaternary Science Reviews, Elsevier, 2019, 212, pp.199-205. �10.1016/j.quascirev.2019.03.001�. �hal-

02083745�

Fire as a motor of rapid environmental degradation during the earliest peopling of Malta 7500 years ago

N. Marriner ^{a , *} , D. Kaniewski ^{b , c , d} , T. Gambin ^e , B. Gambin ^f , B. Vanni ere ^a , C. Morhange ^g , M. Djamali ^h , K. Tachikawa ^g , V. Robin ⁱ , D. Rius ^a , E. Bard ^g

aCNRS, Laboratoire Chrono-Environnement UMR 6249, MSHE Ledoux, USR 3124, Universite de Bourgogne-Franche-Comte, UFR ST, 16 Route de Gray, 25030, Besançon, France

bUniversite Paul Sabatier-Toulouse 3, EcoLab (Laboratoire d’Ecologie Fonctionnelle et Environnement), B^atiment 4R1, 118 Route de Narbonne, 31062, Toulouse cedex 9, France

cCNRS, EcoLab (Laboratoire d’Ecologie Fonctionnelle et Environnement), 31062, Toulouse cedex 9, France

dInstitut Universitaire de France, Secteur Biologie-Medecine-Sante, 103 boulevard Saint Michel, 75005, Paris, France

eDepartment of Classics&Archaeology, University of Malta, Msida MSD, 2080, Malta

fInstitute of Earth Systems, University of Malta, Msida MSD, 2080, Malta

gAix Marseille Univ, CNRS, IRD, INRA, Coll France, CEREGE, Aix-en-Provence, France

hAix Marseille Univ, Univ Avignon, CNRS, IRD, UMR 7263 IMBE, Aix-en-Provence, France

iInterdisciplinary Laboratory for Continental Environments, CNRS, University of Lorraine, France

Article history:

Received 4 December 2018 Received inrevisedform20February2019 Accepted 1 March 2019

Keywords:

Fire Palaeoecology Human impacts Neolithic Holocene Islands Malta Mediterranean

a b s t r a c t

The Holocene colonisation of islands by humans has invariably led to deep-seated changes in landscape dynamics and ecology. In particular, burning was a management tool commonly used by prehistoric societies and it acted as a major driver of environmental change, particularly from the Neolithic onwards.

To assess the role of early human impacts (e.g. livestock grazing, forest clearance and the cultivation of marginal land) in shaping

“

pristine

”

island landscapes, we here present a 350-year record of

fi

re history and erosion from Malta, straddling the earliest peopling of the island. We show that recurrent anthro- pogenic burning related to Neolithic agro-pastoral practices began ~7500 years ago, with well-defined

fi

re-return intervals (FRI) of 15

e

20 years that engendered erosion and rapid environmental degrada- tion. As early as the Neolithic, this study implies that, in sensitive insular contexts, just a few generations of human activities could rapidly degrade natural islandscapes.

1. Introduction

The arrival of even small human populations to uninhabited lands can lead to permanent vegetation shifts, extended biomass burning and increased erosion (Butzer, 2005; Rull et al., 2015).

Islands, in particular, are extremely sensitive to these widespread ecological transformations (Burjachs et al., 2017), especially when fi re is used as a tool of landscape modi fi cation and management (Jouffroy-Bapicot et al., 2016). Early human societies used fi re in technology, social organisation, subsistence, to manipulate the

environment, for clearing occupation areas, and even in warfare (Pausas and Keeley, 2009; Glikson, 2013). With a surface area of just 246 km

²

, Malta constitutes a unique laboratory to explore the ecological dimensions of early human impacts - and associated environmental tipping points - in a Mediterranean island context (Guilaine, 2003; Broodbank, 2013; French et al., 2018). The island has a long and complex history of human occupation, stretching back at least ~7500 years cal. BP (Cassar, 1997). These early settlers formed communities and introduced agriculture (Bonanno, 2011).

They were able to manage the transfer of animal stocks and domesticated plants that necessitated robust sea-craft and navi- gational skills, as well as bring knowledge of managing “ founding stocks ” of animals (Bonanno, 2011). By 4100 BCE, the island had developed a rich Neolithic culture, most enduringly expressed by

*Corresponding author.

E-mail address:nick.marriner@univ-fcomte.fr(N. Marriner).

https://doi.org/10.1016/j.quascirev.2019.03.001

its megalithic architecture, unparalleled elsewhere in the Medi- terranean at this time (Evans, 1971; Vella, 2013).

Here, we couple high-resolution palaeoecological and geochemistry data to provide a new perspective on human impacts and ecosystem dynamics in Mediterranean island contexts, at the time of Malta's earliest peopling. Our study focuses on a fi ne- grained sediment archive from the Marsa fl oodplain, Malta's largest catchment area that drains ~20% of the island (Fig. 1). These new records of regional land-use supplement site-speci fi c archae- ological data, which, at present, are extremely sparse for this key, yet poorly understood, period of Malta's prehistory. In particular, Malta's small island status e it is a mere 27 km long e allows us to probe and better understand the possible impacts of Neolithic so- cieties on a “ pristine ” Mediterranean environment, within the context of wider debate on the Early Anthropocene Hypothesis (Ruddiman, 2007) and human-driven changes in Mediterranean fi re regimes during the early Neolithic (Vanni ere et al., 2016).

2. Methods

A continuous sediment core was extracted from the Marsa fl oodplain in 2011 (Fig. 1; Figs. SM1-2), using a mechanised per- cussion coring system. The fl oodplain lies at the head of a well- protected coastal ria. The Marsa catchment is Malta's largest, with a surface area of 50 km

²

, draining ~20% of the island. The geology of the watershed is dominated by limestones of Miocene age (Pedley et al., 1976) that are dissected by an intricate system of widens, short Mediterranean-type watercourses that carry water and sediment during the wet season and fl ash- fl ood events. Today, the Marsa watershed is heavily arti fi cialised, with a shallow soil cover (20 e 60 cm) made-up of terra rosa, xerorendzinas and car- bonate raw soils (Lang, 1960). The soils are easily eroded under a climatic regime of long dry summers and a wet season in which rain frequently falls in heavy showers. At present, rainfall averages 529.6 mm per annum and is characterised by a marked seasonal distribution, with ~70% of the total falling between October to March (Chetcuti et al., 1992). In many areas, the valley sides and fl oors have been eroded to bare rock. Similar to other islands and

coastal areas in the Mediterranean region, the vegetation of Malta is in fl uenced by the intense heat and low precipitation experienced during the summer months, as well as the human activity that has increased in the area over the last millennia (Grove and Rackham,

Fig. 1.Location map of Malta and the core site on the Marsafloodplain.

Fig. 2.(A) Charcoal Accumulation Rates (CHAR) and (B) Background Charcoal Accu- mulation Rates (BCHAR) and Fire-Return Intervals (FRI) from the Marsafloodplain, for the period 7600 to 7250 cal years BP. The encircled crosses denote statistically-defined fire events. Fire was intentionally used to create open, matorral vegetation commu- nities, conducive to both grazing and cultivation.

2001; Carroll et al., 2012; Gambin et al., 2016). The three main semi- natural vegetation types presently found on the island are maquis (Ceratonia siliqua, Olea europaea, Pistacia lentiscus, Rhamnus ala- ternus and R. oleoides), garrigue (Thymbra capitata, Erica multi fl ora, Euphorbia melitensis, Teucrium fruticans and Anthyllis hermanniae) and steppe (Lygeum spartum (clay slopes) Hyparrhenia hirta, Andropogon distachyus, Brachypodium retusum, Stipa capensis, Aegilops geniculata, Carlina involucrata, Notobasis syriaca, Galactites tomentosa, Asphodelus aestivuus, and Urginea pancration); along with some smaller community types, such as woodland and coastal wetlands, as well as freshwater, sand dune and rocky habitats which are all important for the rare endemic species that can be found within them (Schembri, 1997).

The core was described, wrapped and labelled in the fi eld, then exported to the CEREGE geoscience facility in Aix-en-Provence, where core sections were longitudinally split. We performed high-resolution measurements (0.5 cm) for Fe, Ti and Ca using an XRF core scanner (ITRAX, Cox Analytical Systems) with a Chromium

X-ray source (35 kV, 35 mA) and a 30 s count time. On the basis of the core's radiocarbon chronology, we pinpointed the period around 7500 cal years BP (Fig. SM3; Table SM1) that, according to the archaeological record, corresponds to the earliest peopling of Malta. The core section in question was sampled at continuous 1- cm intervals (n ¼ 141 samples), with each sample representing a time period of between 1 and 4 years. This exceptionally short temporal resolution is ideal for resolving rapid human impacts, management techniques and landscape degradation at the onset of Malta's Neolithic history. Variations in the in fl ux of charcoal par- ticles were used to provide the primary record of past fi res. 1-cm

³

samples were prepared using classic charcoal preparation tech- niques (Vanni ere et al., 2008). Samples were gently washed through a 150 m m sieve. Charcoal fragments were subsequently extracted - in water and using a pipette in order to avoid breakage - under a binocular microscope and placed in a petri dish. Digital photographs and the image-processing computer program ImageJ were used to precisely quantify the surface area of charcoal

Fig. 3.(A) Mean aspect ratios (AR) of charcoal fragments from the Marsa record. The red line denotes the mean AR score and the upper and lower boundaries of the grey envelope the 25th and 75th percentiles. (B) AR variance scores plotted against the charcoal number. (C) Boxplots of the AR variance between 7600 and 7360 yrs BP and 7360-7250 years BP.

(D) Minimum and maximum mean AR scores. (For interpretation of the references to colour in thisfigure legend, the reader is referred to the Web version of this article.)

fragments per 1 cm

³

of sediment aggregate. All charcoal data are reported in mm

²

/cm

³

of sediment (Fig. SM4). There was a robust correlation between the surface area of charcoal samples and the number of pieces in each chronostratigraphic level (r

²

¼ 0.99), which supports our methodology based on computer-aided image processing. Peaks in charcoal are assumed to represent fi re epi- sodes that occurred within or near the Marsa watershed (i.e. at the landscape scale). The image-processing computer program ImageJ was used to calculate charcoal Aspect Ratios (AR) from digital photographs (AR ¼ charcoal [Major Axis]/[Minor Axis]). We ana- lysed 6888 charcoal fragments (Fig. SM5). AR values ranged from 1 to 15.314 (Fig. 2).

All data were analysed using CHAR Analysis, PAST (version 2.17c) and XL-Stat2017. The charcoal concentrations were converted into charcoal accumulation rates (CHAR in mm

²

/cm

²

/yr

¹

) based on the sedimentation rate estimated from the age-depth model. The CHAR record was resampled and normalised to a timescale of 5 yr cm

¹

, corresponding to the mean sedimentation rate of the record (5 yr/

cm

¹

). The data were subsequently log-transformed to homogenise variance (Figs. 2 e 4). The Background component (BCHAR) was estimated with a local polynomial (100-yr moving window), allowing the calculation of the Peak Component (Cpeak ¼ (log)CHAR- (log)BCHAR). Fire episodes were detected using a threshold within the range of values with the lowest sensitivity to the number of peaks detected, derived by plotting a frequency distribution histo- gram of the peak component. The fi re frequency was smoothed using a 70-yr moving window. The process is fully described in the literature (Higuera et al., 2009). The peak component ( fi re-episode detection) is indicated on a linear age-scale.

Periodicities in the CHAR, Fe and Ti/Ca signals were investigated using a sinusoidal regression (phase Free) to determine the long- term trends and further examined using a wavelet analysis (wavelet transform) with Morlet as the basis function, and spectral analysis (Fig. 5).

3. Results

The charcoal record from Marsa supports the occurrence of

extensive human-induced ecological changes and burning of vegetation stands at the time of the earliest peopling of Malta, around 7500 cal years BP (Fig. 2). Charcoal volumes ranged from 0 to 63.8 mm

²

/cm

³

(739 pieces/mm

²

/cm

³

), with a mean of 3.95 mm

²

/cm

³

(49.75 pieces/mm

²

/cm

³

) and a median of 1 mm

²

/ cm

³

(13 pieces/mm

²

/cm

³

). All charcoal volumes were converted into Charcoal Accumulation Rates (CHAR) and then analysed using the method of charcoal signal decomposition developed by Higuera et al. (2009). Statistical analyses detected eleven fi re episodes be- tween 7600 and 7250 cal years BP, with a maximum frequency of 3.67 and a minimum of 0.01 fi re episodes every 70 yr

¹

. The early peopling of the Marsa watershed is expressed by four well-de fi ned in fl uxes of charcoal: (1) a fi rst, more minor, peak (CHAR ~3.5 mm

²

/ cm

²

/yr

¹

) centred on ~7535 cal years BP, followed by a second in fl ection of similar amplitude ~15 years later; and (2) a third pronounced and two-pronged peak spanning ~7490 to 7465 cal years BP (maximum CHAR ~16 mm

²

/cm

²

/yr

¹

). Four fi re episodes attest to close and regular fi re-return intervals (FRI) of ~15 years, beginning ~7535 cal years BP. Two fi re episodes are recorded between ~7390 and 7375 cal years BP, with a further fi ve at ~7335, 7315, 7300, 7280 and 7260 cal years BP.

The landscape dimensions of early anthropogenic fi re activity were explored using high-resolution XRF data. Speci fi cally, Fe counts and Ti/Ca ratios were employed as proxies for watershed soil erosion during the early peopling of Malta, between 7600 and 7250 cal years BP. These geochemical tracers have been shown to be valuable tools in tracking erosion processes (Brisset et al., 2013). To be consistent with the CHAR record, the geochemical data were resampled at regular 5-year intervals and z-score transformed (Fig. 4).

For the period 7600 to 7350 cal years BP, concurrent with the human-induced fi re episodes at ~7535, 7520, 7490 and 7470 cal. BP, we observe a major change in the geochemical regime. These data suggest that, as the soils lost protection from the vegetation cover, they were rapidly eroded by runoff, with pronounced erosion pulses centred on ~7530, 7515, 7490 and 7470 cal years BP (Fig. 4).

For the period ~7600 to 7450 cal years BP, we observe a good cor- relation between the 5-yr CHAR record and the geochemical tracers of erosion (Ti/Ca, r ¼ 0.79; Fe, r ¼ 0.7), which attests to a near instantaneous response of Malta's fragile island landscape to vegetation clearance by burning. A third pulse in Fe and Ti/Ca values ~7425 cal years BP, not accompanied by a peak in charcoal concentrations, implies that erosion continued for several decades in the absence of a well-developed vegetation cover.

4. Discussion

Within the Marsa catchment, our data point to extensive human-induced biomass burning (BCHAR) and an unprecedented increase in fi re frequency, probably driven by Malta's earliest col- onists. We assume that this was done with the aim of clearing the natural landscape for livestock grazing and the cultivation of land, as attest palaeoecological records from elsewhere in the central Mediterranean (Colombaroli et al., 2008; Colombaroli and Tinner, 2013). After ~7470 cal years BP, a FRI of 80 years evokes the possi- bility of either an occupation hiatus or a decline in the need to burn surrounding areas. The end of this hiatus is marked by two fi re episodes between ~7390 and 7375 cal years BP. These events are followed by a further possible break in occupation of ~40 years.

More sustained activity in the area is characterised by fi ve regularly-spaced fi re episodes at ~7335, 7315, 7300, 7280 and 7260 cal years BP. These human-induced palaeo fi re cycles of 15 e 20 years mesh tightly with observed mean fi re-return intervals in present human-modi fi ed ecosystems of the western Mediterra- nean (Mouillot et al., 2003; Pausas and Fern andez-Mu~ noz, 2012).

Fig. 4.Geochemical indicators of erosion, namely z-transformed Ti/Ca (A) and Fe (B).

These are plotted against the 5-yr CHAR (A) and 70 yr-1 Peak Frequency (B) on the left.

The encircled crosses denote statistically-definedfire events.

Such cycles are consistent with time-dependent shrub-vegetation growth and the accumulation of dry biomass from the understorey vegetation of these ecosystems, reaching maturity every 15 e 20 years.

Broadly, the CHAR record between 7250 and 7600 cal years BP indicates that the regular cyclicity of Marsa's palaeo fi re regime was driven by intensifying Neolithic land-use. Climate data from nearby Sicily suggest that this period was relatively “ dry ” , which could have accentuated human-induced fi re activity through an increase in dry biomass (Tinner et al., 2009; Magny et al., 2011). Current archaeological data suggest that Malta's early settlers probably originated from Sicily. These populations have been identi fi ed through the distinctly decorated pottery known as Stentinello Ware (Sagona, 2015). Whether due to push factors caused by de- mographic pressures in their homeland and/or due to the pull factor of new unexploited land, a group (or groups) of persons probably undertook the sea-crossing of just over 80 km. They brought with them all the necessary personnel, tools, technological knowledge, seeds and livestock needed to establish a community (or communities) based on agriculture and animal husbandry (Bonanno, 2011). Archaeological layers from this period of early

human occupation on Malta show that these early settlers main- tained socio-cultural ties with Sicily (Vella, 2008). Such contact would have provided access to a “ support network ” that could, should the need arise, provide much needed replenishments to the fl edgling colony. In turn, this would have led to a scenario in which agriculture e and hence the need for land clearance and manage- ment e could be more easily sustained.

In the main, archaeological evidence for the earliest phases of human occupation on Malta has come from two sites: G ħ ar Dalam, a cave just off St George's Bay in the south east of Malta and Skorba, an inland site in the central-northern part of the island. Impor- tantly, some of the pottery found in these early levels is, despite its close semblance to that originating from Sicily, made of local clay, which is easily sourced in various parts of the island (Tanasi and Vella, 2011). The signi fi cance of this must not be underestimated.

Local clay was being harvested, shaped and decorated in particular

styles but it also had to be fi red. Nothing is known about the type of

kilns used during the early Neolithic period of Malta, as none have

yet been discovered. However, with the right techniques and

knowledge, pottery can also be produced using pit- fi ring (Maggetti

et al., 2011). Whatever the case, the production of local ware from

Fig. 5.Wavelet analyses (A), periodograms (B) and sinusoidalfits (C) of the CHAR, TI/Ca and Fe datasets. These underscore a lag in the landscape response to vegetation burning.

an emerging and expanding community over the period of study must have contributed somewhat to the exploitation of existing trees and vegetation cover, as evoked by recent palynological studies (Carroll et al., 2012; Djamali et al., 2013).

Experimental studies have demonstrated that charcoal mor- phometrics can be used to trace the type of fuel burnt during fi re events (Jensen et al., 2007). In this instance, we used digital image treatment to calculate charcoal Aspect Ratios (AR, i.e. charcoal [Major Axis]/[Minor Axis]) in order to probe trends in the fuel types used by Malta's earliest Neolithic colonists. Crawford and Belcher (2014), for instance, have shown that wood and leaves generally generate ARs < 3, whereas grass tends to produce charcoal with ARs

> 3. We analysed 6888 charcoal fragments in 141 chronostrati- graphic levels (Fig. 3); the maximum AR value was 15.314 and the minimum was 1. Palynological results from the Burmarrad catch- ment evoke a transition from “ natural ” to “ human-modi fi ed ” fuel sources, essentially characterised by the burning of forest stands before 7360 cal yrs BP in contrast to a mixed pyrophytic shrub- grassland, comprising Asphodelus, Erica, Quercus ilex, Pinus, Pista- cia and Sanguisorba, after this time (Djamali et al., 2013). In the Marsa record, overall charcoal in fl ux between 7600 and 7360 cal yrs BP was 76% greater than in the later phase 7360 to 7250 cal yrs BP. Lower aspect ratio variance in the former period suggests that Malta's pioneer colonists burned a relatively narrow range of fuel stocks (i.e. tree stands, Fig. 3).

For the period 7350 to 7250 cal years BP, the charcoal data show well-de fi ned fi re-return intervals of 15 e 20 years. These results point to the establishment of regular patterned burning related to Neolithic practices in the Marsa catchment. Following the initial landscape disturbances around 7500 cal years BP, erosion pulses slightly lag behind fi re episodes (Fig. 5). The wavelet transforms (basis function Morlet) and spectral analyses both display two main periodicities of 175-yr and 50-yr in the fi re regime. Using the major 175-yr cycle as a reference period for the long-term trends (Fig. SM6), the sinusoidal regression (phase Free) depicts a clear lag of 10 years between peaks in fi re and erosion (P < 0.001), sug- gesting that burning activities played a key role in the degradation of soils.

Our data imply that early Neolithic farming and herding activ- ities on Malta developed in coastal areas and lowlands, much like Sicily at this time (Natali and Forgia, 2018). Although some recent publications do cover environmental and agricultural matters (Bonanno and Vella, 2015) such data remains sparse and relatively unexplored. Of notable interest are ceramic zoomorphic handle fragments, depicting bovine heads, from both G ħ ar Dalam and Skorba (Sagona, 2015). Nonetheless, at present, there are no pub- lished archaeological data pertaining to the type of agriculture adopted by Malta's fi rst human populations. This constitutes an important avenue for future research.

5. Conclusion

Global change has sharpened focus on the vulnerability of Mediterranean islands to past, present and future climate modi fi - cations (Vogiatzakis et al., 2016). Profound human impacts and climate risk are notable challenges to small islands, such as the Maltese archipelago, because their pressure absorption ability is lower than for continents. Because Malta is bound by its insularity, it is a particularly relevant unit of analysis to probe how past environmental changes and intense human pressures have impacted islandscapes and ecological systems, and contextualise future global change in Mediterranean island settings (e.g.

McLaughlin et al., 2018).

The high-temporal resolution of data obtained from the Marsa fl oodplain allows us to elucidate a heavy human imprint that

closely mirrors early human occupation of the island of Malta.

These new results provide evidence for deep-seated modi fi cation of Malta's landscapes at the time of the earliest peopling of the island, with wider implications for our understanding of anthropogenic disturbances in “ pristine ” island contexts, as well as ecological tipping points (Reyer et al., 2015).

Acknowledgements

Support was provided by the ANR project “ PALEOMED ” and the Institut Universitaire de France, “ CLIMSORIENT ” project. This work is a contribution to Labex OT-Med (n

ANR-11-LABX-0061). We wish to thank Marta Garcia for assistance with the XRF measurements.

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.quascirev.2019.03.001.

References

Bonanno, A., 2011. The lure of the islands: Malta's first Neolithic colonisers. In:

Phoca-Cosmetatou, N. (Ed.), The First Mediterranean Islanders: Initial Occupa- tion and Survival Strategies. Oxbox Books, Oxford, pp. 145e155.

Bonanno,A.,Vella,N.,2015.Tas-Silg,Marsaxlokk(Malta)IArchaeological Excava-tions Conducted by the University of Malta, 1996-2005. Peeters Publishers,

Leuven.

Brisset, E., Miramont, C., Guiter, F., Anthony, E.J., Tachikawa, K., Poulenard, J., Arnaud, F., Delhon, C., Meunier, J.-D., Bard, E., Sumera, F., 2013. Non-reversible geosystem destabilisation at 4200 cal. BP: sedimentological, geochemical and botanical markers of soil erosion recorded in a Mediterranean alpine lake.

Holocene 31, 1863e1874.

Broodbank, C., 2013. The Making of the Middle Sea - A History of the Mediterranean fromtheBeginningtotheEmergenceoftheClassicalWorld.Thames & Hudson, London.

Burjachs, F., Perez-Obiol, R., Picornell-Gelabert, L., Revelles, J., Servera-Vives, G., Exposito, I., Yll, E.-I., 2017. Overview of environmental changes and human colonization in the Balearic Islands (Western Mediterranean) and their impacts on vegetation composition during the Holocene. J. Archaeol. Sci.: Report 12, 845e859.

Butzer, K.W., 2005. Environmental history in the Mediterranean world: cross- disciplinary investigation of cause-and-effect for degradation and soil erosion.

J. Archaeol. Sci. 32, 1773e1800.

Carroll, F.A., Hunt, C.O., Schembri, P.J., Bonanno, A., 2012. Holocene climate change,vegetation history and human impact in the Central Mediterranean: evidence

from the Maltese Islands. Quat. Sci. Rev. 52, 24e40.

Cassar, L.F., 1997. Settlement patterns in the Maltese Islands: from early colonization to pre-independence. Geojournal 41, 137e144.

Chetcuti, D., Buhagiar, A., Schembri, P.J., Ventura, F., 1992. The Climate of the MalteseIslands: a Review. Malta University Press, Msida, Malta).

Colombaroli, D., Tinner, W., 2013. Determining the long-term changes in biodiver-sity and provisioning services along a transect from Central Europe to the

Mediterranean. Holocene 23, 1625e1634.

Colombaroli, D., Vanniere, B., Emmanuel, C., Magny, M., Tinner, W., 2008. Fire- vegetation interactions during the mesolithic-neolithic transition at lago del-l'Accesa, Tuscany, Italy. Holocene 18, 679e692.

Crawford,A.J., Belcher, C.M., 2014. Charcoal morphometry for paleoecological analysis:theeffectsoffueltypeandtransportationonmorphologicalparam- eters. Applications in Plant Sciences 2, 1400004.

Djamali, M., Gambin, B., Marriner, N., Andrieu-Ponel, V., Gambin, T., Gandouin, E., Medail, F., Pavon, D., Ponel, P., Morhange, C., 2013. Vegetation dynamics during

the early to mid-Holocene transition in NW Malta, human impact versus cli- matic forcing. Veg. Hist. Archaeobotany 22, 367e380.

Evans, J.D., 1971. The Prehistoric Antiquities of the Maltese Islands: A Survey. Ath- lone Press, London.

French, C., Taylor, S., McLaughlin, R., Cresswell, A., Kinnaird, T., Sanderson, D., Stoddart, S., Malone, C., 2018. A neolithic palaeo-catena for the Xagħra upper

coralline limestone plateau of Gozo, Malta, and its implications for past soil development and land use. Catena 171, 337e358.

Gambin, B., Andrieu-Ponel, V., Medail, F., Marriner, N., Peyron, O., Montade, V., Gambin, T., Morhange, C., Belkacem, D., Djamali, M., 2016. 7300 years of vegetation history and climate for NW Malta: a Holocene perspective. Clim. Past 12, 273e297.

Glikson, A., 2013. Fire and human evolution: the deep-time blueprints of the Anthropocene. Anthropocene 3, 89e92.

Grove, A.T., Rackham, O., 2001. The Nature of Mediterranean Europe: an EcologicalHistory. Yale University Press, London.

Guilaine, J., 2003. De la vague a la tombe : Metamorphoses en Mediterranee (8000- 2000. avant J.C.). Seuil, Paris.

Higuera, P.E., Brubaker, L.B., Anderson, P.M., Hu, F.S., Brown, T.A., 2009.

Vegetationmediatedtheimpactsofpostglacialclimatechangeon fireregimesinthe south-central Brooks range, Alaska. Ecol. Monogr. 79, 201e219.

Jensen, K., Lynch, E.A., Calcote, R., Hotchkiss, S.C., 2007. Interpretation of charcoalmorphotypes in sediments from Ferry Lake, Wisconsin, USA: do different plant

fuel sources produce distinctive charcoal morphotypes? Holocene 17, 907e915.

Jouffroy-Bapicot, I., Vanniere, B., Iglesias, V., Debret, M., Delarras, J.-F., 2016. 2000 Years of grazing history and the making of the cretan mountain landscape, Greece. PLoS One 11 e0156875.

Lang, D.M., 1960. Soils of Malta and Gozo. Her Majesty's Stationary Office, London.

Maggetti, M., Neururer, C., Ramseyer, D., 2011. Temperature evolution inside a pot during experimental surface (bonfire) firing. Appl. Clay Sci. 53, 500e508.

Magny, M., Vanniere, B., Calo, C., Millet, L., Leroux, A., Peyron, O., Zanchetta, G., La Mantia,T.,Tinner,W.,2011.Holocenehydrologicalchangesinsouth-western Mediterraneanasrecordedbylake-levelfluctuationsatLagoPreola,acoastal lakeinsouthernSicily,Italy.Quat.Sci.Rev.30,2459e2475.

McLaughlin, T.R., Stoddart, S., Malone, C., 2018. Island risks and the resilience of a prehistoric civilization. World Archaeol. https://doi.org/10.1080/00438243.

2018.1515656.

Mouillot, F., Ratte, J.P., Joffre, R., Moreno, J.M., Rambal, S., 2003. Some determinants of the spatio-temporal fire cycle in a Mediterranean landscape (Corsica, France).

Landsc. Ecol. 18, 665e674.

Natali, E., Forgia, V., 2018. The beginning of the neolithic in southern Italy and sicily.Quat. Int. 470, 253e269.

Pausas, J.G., Fernandez-Mu~noz, S., 2012. Fire regime changes in the Western Med- iterranean Basin: from fuel-limited to drought-driven fire regime. Clim. Change 110, 215e226.

Pausas, J.G., Keeley, J.E., 2009. A burning story: the role of fire in the history of life.

Bioscience 59, 593e601.

Pedley, H.M., House, M.R., Waugh, B., 1976. The geology of Malta and Gozo. Proc.

Geologists' Assoc. 87, 325e341.

Reyer, C.P.O., Brouwers, N., Rammig, A., Brook, B.W., Epila, J., Grant, R.F.,Holmgren, M., Langerwisch, F., Leuzinger, S., Lucht, W., Medlyn, B., Pfeifer, M.,

Steinkamp, J., Vanderwel, M.C., Verbeeck, H., Villela, D.M., 2015. Forest resil- ience and tipping points at different spatio-temporal scales: approaches and challenges. J. Ecol. 103, 5e15.

Ruddiman, W.F.,2007.Theearlyanthropogenichypothesis:challengesandre- sponses. Rev. Geophys. 45, RG4001.

Rull,V.,Ca~nellas-Bolta,N.,Margalef,O.,Saez,A.,Pla-Rabes,S.,Giralt,S.,2015.

LateHolocene vegetation dynamics and deforestation in Rano Aroi: implications for Easter Island's ecological and cultural history. Quat. Sci. Rev. 126, 219e226.

Sagona, C., 2015. The Archaeology of Malta: from the Neolithic through the Roman

Period. Cambridge University Press, Cambridge.

Schembri, P.J., 1997. The Maltese Islands: climate, vegetation and landscape. Geo- journal 41, 1e11.

Tanasi, D., Vella, N.C., 2011. Site, Artefacts and Landscape: Prehistoric Borg_ In- Nadur.Malta. Polimetrica, Monza.

Tinner, W., van Leeuwen, J.F.N., Colombaroli, D., Vescovi, E., van der Knaap, W.O.,Henne,P.H.,Pasta,S.,D'Angelo,S.,LaMantia,T.,2009.Holoceneenvironmental

andclimaticchangesatGorgoBasso,acoastallakeinsouthernSicily,Italy.

Quat.Sci.Rev.28,1498e1510.

Vanniere,B.,Colombaroli,D.,Chapron,E.,Leroux,A.,Tinner,W.,Magny,M.,2008.

Climate versus human-driven fire regimes in Mediterranean landscapes: the Holocene record of Lago dell'Accesa (Tuscany, Italy). Quat. Sci. Rev. 27, 1181e1196.

Vanniere, B., Blarquez, O., Rius, D., Doyen, E., Brücher, T., Colombaroli, D., Connor, S., Feurdean, A., Hickler, T., Kaltenrieder, P., Lemmen, C., Leys, B., Massa, C., Olofsson, J., 2016. 7000-year human legacy of elevation-dependent European fire regimes. Quat. Sci. Rev. 132, 206e212.

Vella,C.,2008.EmergingaspectsofinteractionbetweenprehistoricSicilyandMalta from the perspective of lithic tools. In: Bonanno, A., Militello, P. (Eds.), Malta and the Hybleans, the Hybleans in Malta. Koine Archeologica Sapiente Anti- chita, Palermo, pp. 81e93.

Vella, G., 2013. G_gantija: the Oldest Free-Standing Buildings in the World (Heritage Malta, Malta).

Vogiatzakis, I.N., Mannion, A.M., Sarris, D., 2016. Mediterranean island

biodiversityand climate change: the last 10,000 years and the future. Biodivers. Conserv. 25, 2597e2627.